Studies of amplification in rhabdomyosarcoma

为了进一步研究12q13-q14染色体区域在FP RMS中的扩增，我们之前使用了癌症基因组图谱（TCGA）的数据，将我们在RMS中的发现与多形性胶质母细胞瘤、去分化脂肪肉瘤和肺腺癌中该区域的扩增进行了比较。我们对这四种癌症类别中的12q13-q14扩增子的分析显示，扩增的高置信度区域从多形性胶质母细胞瘤的0.5 Mb到肺腺癌的1.6 Mb。对这些扩增区域的比较发现，在四种癌症类别中存在0.2 Mb的重叠，以及FP RMS特异性的0.5 Mb扩增区域的存在。通常扩增的0.2 Mb区域包含15个基因，其中7个基因（包括CDK4）在所有四种癌症类别的扩增子阳性样本中过表达。此外，FP RMS特异性的0.5 Mb区域包含18个基因，其中4个基因在扩增的FP RMS中过表达（NEMP1、NAB2、SHMT2和R3HDM2）。为了进一步研究这一扩增片段对FP RMS的特异性，我们分析了NEMP1、NAB2、SHMT2和R3HDM2在8个FP RMS细胞系中的表达，其中7个细胞系没有12q13-q14扩增子，1个细胞系（RH30）有12q13-q14扩增子。表达研究显示，与其他7种FP RMS细胞系相比，RH30细胞中NAB2和SHMT2 mRNA的表达水平较高，随后的western blot研究显示，RH30细胞中只有SHMT2蛋白水平过表达。作为后一发现的证实，我们对13个FP RMS pdx的研究也显示，与未扩增的10个pdx相比，具有SHMT2扩增的3个pdx中的SHMT2 RNA和蛋白表达更高。为了研究SHMT2在FP RMS细胞中表达变化的后果，我们将几种靶向SHMT2的慢病毒shRNA构建体转导到RH30中，导致SHMT2表达降低。这种SHMT2表达的减少伴随着细胞在低密度下培养时生长下降和菌落生长下降。此外，体外转化活性分析显示，shrna介导的表达减少导致RH30细胞的病灶形成大幅减少。最后，注射到免疫功能低下的小鼠中也显示其致瘤活性显著降低。因此，RH30细胞中SHMT2表达的缺失导致致癌活性的显著降低，这与细胞增殖和/或存活的变化有关。在一组补充实验中，我们将SHMT2的cDNA表达构建体转导到两个FP RMS细胞系（Rh5和RH41）中，没有12q13-q14扩增，将SHMT2的表达提高到与扩增的RH30细胞相似的水平。在这些研究中，SHMT2表达的增加与细胞在低密度下培养时的生长增加和菌落生长增加有关。此外，对致瘤活性的分析显示，SHMT2表达的增加导致体外灶形成和体内致瘤活性的增加。在最后一组实验中，我们检测了直接或间接抑制shmt2相关信号的药物。SHMT2的直接抑制剂SHIN1在我们的原生和转基因FP RMS细胞系中的研究表明，SHIN1的敏感性与SHMT2的扩增和表达成反比。我们将这种直接作用药物在SHMT2高表达情况下的效力降低归因于SHMT2蛋白的高水平结合和隔离药物的能力，从而降低了药物效力。接下来，我们进一步沿着信号通路寻找一种抑制下游蛋白质的药物，这种蛋白质在不改变蛋白质表达的情况下被激活。由于叶酸通路位于SHMT2的下游，我们选择了叶酸抑制剂培美曲塞。培美曲塞在我们的原生和转基因FP RMS细胞系中的类似分析显示，培美曲塞的敏感性与SHMT2的扩增和表达成正比，这表明培美曲塞可能为具有12q13-q14扩增的FP RMS亚群提供靶向治疗。总之，我们的研究表明，FP RMS中的12q13-q14扩增子包含至少两个驱动癌基因。除了作为扩增子的致癌驱动因素外，这些基因中至少有一个可能是用扩增子进行肿瘤分子定向治疗的有用靶标。

To further study amplification of the 12q13-q14 chromosomal region in FP RMS, we previously used data from The Cancer Genome Atlas (TCGA) to compare our findings in RMS with amplification of this region in glioblastoma multiforme, dedifferentiated liposarcoma, and lung adenocarcinoma. Our analysis of the 12q13-q14 amplicons in these four cancer categories revealed high confidence regions of amplification ranging from 0.5 Mb in glioblastoma multiforme to 1.6 Mb in lung adenocarcinoma. Comparison of these amplified regions identified an overlap of 0.2 Mb across the four cancer categories as well as the presence of a 0.5 Mb amplified region that was specific to FP RMS. The commonly amplified 0.2 Mb region contained 15 genes, of which seven genes (including CDK4) were overexpressed in amplicon-positive samples in all four cancer categories. In addition, the 0.5 Mb region specific to FP RMS contained 18 genes, four of which were overexpressed in amplified FP RMS (NEMP1, NAB2, SHMT2 and R3HDM2). To further study this amplified segment that is specific to FP RMS, we analyzed expression of NEMP1, NAB2, SHMT2 and R3HDM2 in a panel of 8 FP RMS cell lines, consisting of 7 lines without and one line (RH30) with the 12q13-q14 amplicon. Expression studies revealed higher levels of NAB2 and SHMT2 mRNA in RH30 cells compared to the other seven FP RMS lines, and subsequent western blot studies revealed that only SHMT2 was overexpressed at the protein level in RH30 cells. As confirmation of this latter finding, our study of 13 FP RMS PDXs also showed higher SHMT2 RNA and protein expression in the 3 PDXs with SHMT2 amplification compared to the 10 PDXs without amplification. To examine the consequences of changes in SHMT2 expression in FP RMS cells, several lentiviral shRNA constructs targeting SHMT2 were transduced into RH30 and resulted in decreased SHMT2 expression. This decrease in SHMT2 expression was accompanied by decreased growth in bulk culture and decreased growth of colonies when cells were plated at low density. Furthermore, analysis of in vitro transforming activity revealed that the shRNA-mediated decrease in expression results in a substantial decrease in focus formation of RH30 cells. Finally, injection into immunocompromised mice also revealed a substantial decrease in tumorigenic activity. Therefore the loss of SHMT2 expression in RH30 cells causes a substantial decrease in oncogenic activity that is related to changes in cellular proliferation and/or survival. In a complementary set of experiments, we transduced a cDNA expression construct for SHMT2 into two FP RMS lines (Rh5 and RH41) without 12q13-q14 amplification to increase SHMT2 expression to levels similar to those seen in amplified RH30 cells. In these studies, the increased SHMT2 expression was associated with increased growth in bulk culture and increased growth of colonies when the cells were plated at low density. Furthermore, analysis of oncogenic activity revealed that increased SHMT2 expression resulted in an increase in focus formation in vitro and tumorigenic activity in vivo. In a final set of experiments, we examined drugs that directly or indirectly inhibit SHMT2-associated signaling. Studies of SHIN1, a direct inhibitor of SHMT2, in our native and genetically-modified FP RMS lines demonstrated that SHIN1 sensitivity is inversely proportional to SHMT2 amplification and expression. We attribute the decreased potency of this direct-acting agent in the presence of high SHMT2 expression to the ability of high levels of SHMT2 protein to bind and sequester the drug, thereby decreasing drug potency. We next moved further down the signaling pathway to find a drug that inhibits a downstream protein that is activated without changes in protein expression. Since the folate pathway is downstream of SHMT2, we selected the folate inhibitor pemetrexed. Similar analysis of pemetrexed in our native and genetically-modified FP RMS lines revealed that pemetrexed sensitivity is directly proportional to SHMT2 amplification and expression, thus suggesting that pemetrexed may provide a targeted therapy for the subset of FP RMS with 12q13-q14 amplification. In conclusion, our studies demonstrate that the 12q13-q14 amplicon in FP RMS contains at least two driver oncogenes. In addition to functioning as oncogenic drivers of the amplicon, at least one of these genes may be a useful target for molecular-directed therapy of tumors with the amplicon.